Algorithm uses personality profile data to improve communication

Crystal is a web-based service that analyzes online profiles to suggest how to build a relationship with anyone.

Sending that first email to an unknown contact can be intimidating. What kind of person are they? Should the email be conversational or to the point? A new tool called Crystal is now addressing this anxiety, by offering users a way to understand someone’s communication style before approaching them. The web-based service aims to make communication healthier and more productive, by helping users write emails based on the recipient’s unique personality.

Working primarily through LinkedIn, Crystal analyzes online content written by and about the chosen recipient. This information is then fed into an algorithm that matches all the available web content to a given personality type. The more visible a person is online, the more accurately the algorithm will predict their communication style — a confidence score is also provided.

When writing emails, Crystal offers suggestions by supplying whole templates or offering alternative phrases and words. Users will be told if they’ve written too much, whether to avoid unnecessary sign-offs, or even to include an emoji or two. The aim is to build an empathetic relationship; writing in a way that appeals to the recipient’s personality. Suggestions are not limited to email — Crystal suggests how to approach conversations with a person, down to favorable discussion topics and the way to greet them. Notably, Crystal alleviates approach anxiety, especially in the case where a user is looking to pitch a sale, ensuring poor communication doesn’t get in the way of business.

Could the language of online adverts change depending on the reader’s personality profile?